Understanding flow through an orifice (FIV/AIV)

[vovozim]

Hi Guys.

I have a re-pressurisation line around an actuated valve which requires a low dP to open.

how do I consider velocity of the gas downstream of the RO?

I will get choked flow, this condition will last a fair while as the volume downstream is fairly large.

I have sized an RO 18.3mm based on a mass flow rate (~6kg/s) to avoid AIV (Accoustic Induced Vibration), but my colleagues and I have had a robust discussion on how to calculate downstream velocity to calculate FIV.

The velocity I get assuming atmospheric conditions initially downstream (DN50 Sch160 line) is a potential velocity well above sonic. Clearly this is not possible. I have sized my RO using a vetted calculation sheet based on upstream conditions and full (13,500kPa) drop to atmospheric, so it should be ok. How then is my velocity so high downstream?

At the moment it sounds like I will get choked flow through the orifice and then a second choke albeit only for a short amount of time in the pipe itself.

Heres hoping I am missing something simple. Also what condition will provide the worst ρV² value when considering re-pressurisation?

Thanks

 

[Latexman]

I'm a little lost without a dimensioned drawing/sketch, but successive multiple chokes is common. I see them with a PSV with a tailpipe. Choked in the PSV nozzle and at the end of the tailpipe.

Good luck,
Latexman

Technically, the glass is always full - 1/2 air and 1/2 water.

 

[Latexman]

The velocity I get assuming atmospheric conditions initially downstream (DN50 Sch160 line) is a potential velocity well above sonic. Clearly this is not possible.

Correct, that is not possible (without an isentropic concerging/diverging nozzle)

I have sized my RO using a vetted calculation sheet based on upstream conditions and full (13,500kPa) drop to atmospheric, so it should be ok. How then is my velocity so high downstream?

Because it is! You looked at the orifice alone. You need a more sophisticated tool to look at the orifice and downstream piping together. I would guess there will be choked flow where the downstream pipe, that the orifice is in, ties into the main pipe. This plus frictional loses will put backpressure on the downstream side of the orifice. If this backpressure causes the orifice pressure ratio to exceed the critical pressure ratio, there will be subsonic flow through the orifice. If this backpressure does NOT cause the orifice pressure ratio to exceed the critical pressure ratio, there will be sonic flow through the orifice. (I suspect the latter will be what happens, but I'm flying somewhat blind without some more details.)

Good luck,
Latexman

Technically, the glass is always full - 1/2 air and 1/2 water.

 

[vovozim]

Thanks Latexman,

We are getting a chemical engineer to complete a dynamic simulation using HYSYS so that should give us an answer.

However, for my own understanding and because this has frustrated me so much.

Where does the second choke occur? At the fitting with the highest K value (e.g. the outlet of the DN50 line into the DN250 line) or does it occur at some calculable distance from the original choke (Orifice plate) or does it occur multiple times along the 400mm section between the orifice and the outlet?

What would be the upstream conditions for this second choke? Considering your mass flow would increase as the pressure increases between the first and second chokes (thus increasing density between the chokes) up to a point where potentially the flow through the orifice is no longer choked and the downstream choke is now the governing condition.

If I was to take the RO out would I then just consider the entire bypass line as an RO in a DN250 line, if so where does the choke occur? At the exit of the bypass line?

I will try pop a sketch up when I get a chance.

Cheers

 

[Latexman]

I look forward to your dimensioned sketch.

Good luck,
Latexman

Technically, the glass is always full - 1/2 air and 1/2 water.

 

[vovozim]

Sketch attached.

 

[Latexman]

That's exactly how I envisioned it, so no need to change my comments.

Why not just put a time delay on the permissive?

Good luck,
Latexman

Technically, the glass is always full - 1/2 air and 1/2 water.

 

[apetri]

I agree with Latexman,
for an accurate solution if you know (as you should) the volumes (downstream RO) you can simply integrate a pressurization curve solving for example 10 points from p initial to p final,
I do this in Excel with PRODE PROPERTIES to simulate RO and related piping,
a run takes a few seconds and I can evaluate results (pressurization time and pressures vs. time) for different RO and piping...